Biomarker Discovery for Meta-Classification of Melanoma Metastatic Progression Using Transfer Learning

Abstract

Melanoma is considered to be the most serious and aggressive type of skin cancer, and metastasis appears to be the most important factor in its prognosis. Herein, we developed a transfer learning-based biomarker discovery model that could aid in the diagnosis and prognosis of this disease. After applying it to the ensemble machine learning model, results revealed that the genes found were consistent with those found using other methodologies previously applied to the same TCGA (The Cancer Genome Atlas) data set. Further novel biomarkers were also found. Our ensemble model achieved an AUC of 0.9861, an accuracy of 91.05, and an F1 score of 90.60 using an independent validation data set. This study was able to identify potential genes for diagnostic classification (C7 and GRIK5) and diagnostic and prognostic biomarkers (S100A7, S100A7, KRT14, KRT17, KRT6B, KRTDAP, SERPINB4, TSHR, PVRL4, WFDC5, IL20RB) in melanoma. The results show the utility of a transfer learning approach for biomarker discovery in melanoma.

Neurotransmitter receptor-related gene signature as potential prognostic and therapeutic biomarkers in colorectal cancer

“…And Huang et al confirmed that GABRD receptors indicated by T cells directly inhibit CD8 + T cells by participating in signal regulation ( Huang et al, 2022 ). Glutamate ionotropic receptor kainate type subunit 3 (GRIK3) and glutamate receptor ionotropic kainate-5 (GRIK5) are members of the glutamate kainate receptor family and play crucial roles in the neuroactive ligand-receptor interaction pathway ( Fang et al, 2021 ; Minoza et al, 2022 ). There is compelling evidence suggesting that GRIK3 participates in cancer progression.…”

“…Xiao et al found that GRIK3 promotes epithelial-mesenchymal transition in breast cancer cells by regulating SPDEF/CDH1 signaling ( Xiao et al, 2019 ). Furthermore, GRIK5 has been identified as a potential biomarker for melanoma metastatic progression ( Minoza et al, 2022 ). Although limited research exists on the relationship between GRIK5 expression and colorectal cancer (CRC), our study fills this gap by identifying a significant association between high GRIK5 expression and CRC progression.…”

Current knowledge about immunotherapy resistance for melanoma and potential predictive and prognostic biomarkers

“…In addition, the boundary between predictive and prognostic biomarkers is usually unclear, so when people discover a new biomarker, it is impossible to categorize it definitively as predictive or prognosis [ 162 ] . To address these challenges, multi-omic approaches, such as integrating genomics, transcriptomics, and proteomics data, are employed to identify comprehensive biomarker signatures [ 163 , 164 ] . Machine learning algorithms and artificial intelligence are also used to analyze large datasets and uncover novel associations between biomarkers and treatment responses [ 165 , 166 ] .…”

Emerging Prognostic and Predictive Significance of Stress Keratin 17 in HPV-Associated and Non HPV-Associated Human Cancers: A Scoping Review

“…The details of biomarker assessment methodologies are presented in Supplemental Table S2. Studies investigating K17 at the transcriptome level (n = 24) have found K17 to hold negative prognostic value in most cancer entities (endometrial [30,71], esophageal [57], gastric [72], HNSCC [38,42,52], lung cancer [73,74], melanoma [75,76], pancreatic adenocarcinoma [50,51], and renal cell carcinoma [77]). In TNBC, a negative correlation between K17 RNA expression and outcome was found only in a subgroup of patients with invasive ductal carcinoma with large tumors and advanced stage [29].…”

“…The detailed characteristics and key findings of prognostic transcriptome-based studies are presented in Table 4 and Table S4. Thirteen studies (56.5%) were based on data derived from The Cancer Genome Atlas (TCGA) [30,31,49,50,57,71,71,73,75,76,80]. Five studies (21.7%) generated their own transcriptome data [38,42,51,52,79].…”